Liver cancer, mainly hepatocellular carcinoma (HCC), has become a most deadly malignant disease worldwide. So far, pharmaceutic inhibition of major oncogenic pathways has achieved little therapeutic benefit to liver cancer patients. We believe this is due to under- appreciation of the complexity in mechanisms of hepato-oncogenesis. In recent experiments, we and others have identified paradoxically anti-oncogenic effects of classical oncogenic molecules, such as c-Met, EGFR, ?-catenin, Ikkb, Jnk, and Shp2, in the liver. Ablating these molecules in hepatocytes enhanced HCC induced by chemical carcinogen DEN. To test a theory that loss of the oncogenic molecules generates an oncogenic microenvironment that promotes DEN-induced HCC, we have established another mouse HCC model, by transfection of oncogenic ?-catenin (CAT), c-Met (MET) and PIK3CA (PIK), oncoproteins frequently detected in human HCCs. As expected, MET/CAT-driven HCC was aggravated in ?-catenin-deficient liver, due to tumor-promoting factors induced by ?-catenin removal. In contrast, Shp2 deletion dramatically suppressed HCC driven by MET/CAT or MET/PIK, despite a similar pro-tumorigenic environment in Shp2-deficient liver. Based on these novel unanticipated data, we propose a new hypothesis that although removal of Shp2 or ?-catenin generates cell-extrinsic tumorigenic factors in the hepatic environment, the endogenous Shp2 is indispensable for oncogenic signaling in hepatocytes. To test this hypothesis, we propose three specific aims on this project. Aim 1 is to determine the cell-intrinsic role of Shp2 in hepato-oncogenic signaling. Aim 2 is to determine the cell-autonomous effect of ?-catenin in liver tumorigenesis. Aim 3 is to search and identify cell-extrinsic factors induced by loss of the oncoproteins in hepatocytes. The results are expected to be instrumental for design of novel therapeutic strategies for liver cancer by inhibiting both cell-intrinsic oncogenic signals and the secondary environmental factors.